Introduction
There was a workers’ committee that was in charge of the strikes [of 2012]. The workers’ committee members, they became office members in AMCU [Association of Mineworkers and Construction Union]. As an AMCU [leader] you are no longer a workers’ committee member … That was extinct, now you are an office bearer of AMCU. So they brought the norms and values of the workers’ committee into the AMCU atmosphere … Sometime during the strike, we knew that eventually we were going to belong to a union … We are not at AMCU, as I said, not because of a particular ideological common ground. We don't have any common ground with AMCU. As the striking [workers’] committee, [there is] none. It was a matter of compliance that for us to be recognised we need[ed] to belong to a union. Yeah, we joined AMCU.
An AMCU shop steward eloquently captured the emerging dynamics between the independent – or non-union aligned – workers’ committee at Anglo Platinum1 and the new union (AMCU) that they eventually joined and became leaders of. As he signalled in 2013, ‘the relationship between the workers’ committee and AMCU is not that cosy’ (interview, AMCU shop steward).2 Others have indicated that the two are antagonistic towards each other. Prior to the 2012 unprotected strike at Amplats (which followed concurrent and unprecedented unprotected strikes at Amplats and Lonmin), workers engaged with the National Union of Mineworkers (NUM) shop stewards directly.3 The NUM was then the dominant union in the platinum mines and the largest trade union in the country.4 Furthermore, it was a powerhouse in the Congress of South African Trade Unions (COSATU), which forms part of the tripartite alliance with the ruling African National Congress (ANC) and South African Communist Party. Workers began a process that would, to some extent, challenge the hegemony of the ruling party and transform industrial relations in South Africa.
Tensions between trade unions and worker committees, or trade union leadership and rank-and-file mobilisation, have an extensive history and there is a vast literature on this topic at an international level, particularly within the field of Sociology (see, for example, Hyman 1975). Since the publication a century ago of Michels’ (1915) critique of power in organisations, there has been thorough debate regarding whether trade unions are more prone to oligarchic tendencies than political parties. As organisations become larger and institutionalised, according to Michels (1915), they nurture a small minority of leaders who inherently become bureaucrats with specialist decision-making authority. The assumption is that the rank and file would respond by becoming passive and subdued. He calls the supposed fact that power inevitably corrupts leaders the ‘Iron Law of Oligarchy’ (Ibid.).
This article suggests that Goulder's (1955) critique of Michels provides a useful starting point from which to understand the contemporary mineworkers’ movement in South Africa. In particular, it may help explain the relationship between the trade union AMCU and the worker committees which were formed from 2012 at Amplats. Gouldner argued that Michels presented a model which was monolithic and static rather than socially constructed and contextually specific:
It is the ethos of pessimism, rather than compulsions of rigorous analysis, that leads to the assumption that organizational constraints have stacked the deck against democracy. For on the face of it there is every reason to assume that ‘the underlying tendencies which are likely to inhibit the democratic process' are just as likely to impair authoritarian rule. It is only in the light of such a pessimistic pathos that the defeat of democratic values can be assumed to be probable, while their victory is seen as a slender thing, delicately constituted and precariously balanced … the very same evidence to which he called attention could enable us to formulate the very opposite theorem – ‘the iron law of democracy’ … if oligarchical waves repeatedly wash away the bridges of democracy, this eternal recurrence can happen only because men doggedly rebuild them after each inundation. Michels chose to dwell on only one aspect of this process, neglecting to consider this other side. There cannot be an iron law of oligarchy, however, unless there is an iron law of democracy. (Gouldner 1955, 505–506)
This dialectic between oligarchy and democracy, and (to put it another way) between rank and file and top-down control, has remained embedded in the practices of workplaces in South Africa since the independent black trade union movement gained traction following the 1973 Durban strikes. By 1974, the government had established 1700 ‘liaison committees’ made up of worker leaders in an attempt to institutionalise insurgency, though this would, to a significant extent, in fact do quite the opposite. Rising militancy in both communities and the workplaces led the racist white minority-rule apartheid government to establish the Wiehahn Commission (from 1977) – an even more thoroughgoing attempt to institutionalise rank-and-file organising. It created possibilities for certain changes in the working conditions of black people, but nevertheless maintained apartheid's fundamental framework of exclusion (von Holdt 2003, 61). Throughout the 1980s, rank-and-file organisation gave rise to legally recognised trade unions under the Wiehahnian reforms of the apartheid regime.
Formalised leadership within an array of unions across the country sought to maintain workers’ direct control over decision-making processes. Simultaneously, however, they took the lead with regard to workers’ issues and the political trajectory of the trade union movement more broadly, as workers and their unions, in the mid 1980s, became immersed in the anti-apartheid movement. Karl von Holdt (2003) notes that at Highveld Steel under the auspices of the Metal and Allied Workers Union (MAWU) and then later the National Union of Metalworkers of South Africa (NUMSA), tensions arose between shop stewards who used institutional procedures to influence management directly and strike committees who were directly linked to the pulse of the rank and file. Once a ‘guerilla fighter’ (Webster 1984, 81) acting as a mouthpiece of the rank and file, Webster suggests that the shop steward would over time become tame, ‘managing discontent’, as the union became relatively ‘normalized’ into collective bargaining structures (Webster 1985, 235–236, 244, in von Holdt 2003, 7).
In the post-apartheid period, Buhlungu in particular has noted the anti-democratic nature of unions, the tendency for shop stewards to drift away from workers and become part of the bureaucracy, rather than to represent them, as well as the discontent that has resulted from this. He has also called the victory of 1994, whereby the federation COSATU gained great influence as a partner with the ruling ANC, ‘A Paradox of Victory’ (Buhlungu 2010). In keeping with this trend, the NUM, once courageously backing the black oppressed in the struggle for liberation from white apartheid rule (Moodie 2009), became, from the perspective of some activists in the platinum belt, ‘The National Union of Management’ (Moodie 2010).
The formation of independent worker committees and the strikes they helped organise and sustain in the period between 2012 and 2014 were by no means isolated events. They are a reflection of ongoing contestation over union representation at the platinum mines which dates back to at least the early 1990s.5 In the lead-up to the 2012 strike wave, mining companies themselves exhibited some of the most profound inequalities between workers and senior management, despite the long-standing presence of the NUM. Economists at the Labour Research Service, in a survey conducted on mining company CEOs' salaries in 2011, indicated that the average CEO made R20.2 million per year or R55,000 per day (Taal, Patel, and Elsley 2012). This made the demand for R12,500 per month, and R16,070 per month at Amplats, seem like a pittance. Inequality is not in itself an adequate explanation for why mineworkers revolted. At Amplats (as well as Impala and Lonmin) the work process involved conventional rather than mechanised mining. That meant that underground workers, and especially rock drill operators (RDOs), had more power at the workplace since the mines required drillers (and, I would add, other categories of workers) to create the conditions for the removal of platinum. Unlike the gold and coal industries, managers at the platinum mines had also developed a tendency, going back to the 1990s, of engaging directly with worker committees around wage demands (Stewart 2013).6
Dunbar Moodie and Vivienne Ndatshe (1994) indicated that historically, whoever had power in the compounds determined who had control of the mines. With the adoption of the Living Out Allowance after 1994, however, workers have tended to take the allowance and to find their own accommodation, often in the ‘shacklands’ that have proliferated in the Rustenburg region, creating new geographies and dynamics of mineworker settlement.7 As the following case suggests, during the contemporary period at the mines informal organising occurred outside the hostels, next to the workplace and in the changing rooms.
In January 2012, Impala Platinum mine witnessed the beginning of that year's major unprotected strike wave – the subject of Chinguno's article in this issue. Later that month, the unrest had spread beyond the confines of RDOs and virtually the entire workforce had downed tools. They won an increase to R9000 and the strike ended on 3 March. Discontent with the NUM had come to a head at Impala because the union had opposed the popular unprotected strike. By the end of the strike, management had fired 18,000 of the striking mineworkers, which would have ended their union membership. About 11,000 others had resigned as NUM members at Impala by 30 March 2012. Perceived shortcomings of the union at each of the three major platinum mining houses, especially its failure to defend workers subject to dismissal for engaging in unprotected strikes, provided the structural conditions on which informal worker committees took hold first at Impala and then at Lonmin and Amplats respectively.
For socialist activists working closely with the workers in the platinum belt, it appeared that these unprotected strikes of 2012 at the three platinum mines were the pinnacle of resistance. A swift decline in mobilisation was likely to follow as a result of workers’ and their leaders’ decision to join what appeared to be another top-down, authoritarian union. In the midst of and following the relative success of the worker committee at Amplats in pressuring the employer for wage increases, workers decided to join AMCU. The National Executive Committee (NEC) of AMCU swiftly and decisively then sought to crush the worker committee at Amplats. AMCU, perhaps not surprisingly, was opposed to the form of strike (unprotected) that worker committees led and sustained in 2012. It seemed to many observers in 2013, myself included, that under the banner of the new union, unprotected strikes were not an option, and worker militancy would soon perish. Such assumptions conform to a Marxist critique of trade unions as ‘managers of discontent’, limited by their incorporation into bargaining structures and dependent on management support (Cliff and Gluckstein 1986, 21).8
However, while certain aspects of AMCU's approach to the worker committee complied with this prediction, the empirical research provided below also indicates that the core politics which underpinned the militant strikes of 2012 have remained constant over time, despite workers’ decision to join a union and to engage primarily in protected strikes. Hence the 2014 strike united the three largest platinum mines in the world as mineworkers downed tools for five months in what culminated in the longest strike in South African mining history. Put in a different way, this article demonstrates that when the rank and file takes on an insurgent character, the trade union's bureaucratic or official power (at the national, regional and branch level) becomes marginal but only relatively so, as the events reveal. In line with Gouldner's critique of Michels, oligarchy is diminished to the extent that democracy and, I would add, workers’ power flourish. Just as the exclusive nature of the NUM provided the structural basis for new forms of organisations to emerge (worker committees), worker committees created the political space, or at least the possibility, for the flourishing of an Insurgent Trade Union.
After the 2012 unprotected strike at Amplats, the company witnessed in 2013 a struggle over union recognition and also a defensive battle against the company's decision to retrench workers. The leaders of the workers’ committee were absorbed into AMCU structures and Joseph Mathunjwa, president of AMCU – who, in 1999, had formed the union as a result of an internal spat with the NUM – quickly sought to dissolve the committee itself in the Amplats Rustenburg region. While it no longer met formally, the culture that permeated the committees remained in the hearts and minds of rank-and-file workers and amongst the leadership more specifically. One of these leaders thus reflected that they were ‘AMCU by day, workers’ committee by night’ (personal communication, Solomon, 26 May 2013), highlighting the clandestine role of the committee during this phase of transition to the formalised union. I provide an explanation of the origins of the strike at Amplats which was characterised by direct democracy by the rank and file and subsequently demonstrates tensions that emerged when key leaders of the workers’ committee became embedded agents within AMCU.
Origins of the democratic workers’ committee at Amplats
The unprotected strike at Amplats lasted from 11 September 2012 until 18 November 2012 – an impressive period of nearly 10 weeks, longer than both Lonmin and Impala, which had also experienced unprecedented strikes in the platinum industry earlier that year. The immediate roots of the 2012 strike action in Amplats and the parallel formation of informal worker committees outside of established union structures can be traced back to April 2012. This means that the origins of the strike action are not directly linked to the Marikana massacre, which took place on 16 August. Rather, it began when Solomon Mbele, an Amplats worker who is central to the events that follow, had something in mind – something that would in the next several months capture the imagination of Amplats’ entire Rustenburg operation and beyond.
Unlike the wage demands at Impala and Lonmin, which were initiated by RDOs, at Amplats they were initiated by winch drivers. Like many of the workers at Khuseleka – one of the major shafts at Amplats – Solomon was not a member of a union before or during the 2012 strike. He had matriculated (completed high school) in 1999, worked as a truck driver at Goldfields in Carletonville from 2003 until 2009 and arrived at Amplats, where he got a job as a winch driver in January 2010. He was never a member of a union at the mines until January 2013. He and his fellow workers then decided to join AMCU. It was a decision that will be revisited later in this article. Solomon could see that the NUM ‘failed us at Goldfields’.9 He remembers that by the time he came to work in Rustenburg, ‘I was fed up with the NUM. I was fed up with the way they treated the workers’ (interview, Solomon).
Solomon worked night shifts with another winch driver named Anele, who had come to the mines in 1978 and had had his front teeth smashed out by the apartheid police in 1985 while struggling to bring the NUM to the mines. He was a long-standing leader with vast experience – the quintessential insider. At 56, Anele was 22 years older than Solomon. Reflecting on the relationship that they had, Solomon stated that ‘he can't be my friend … about the issues that were happening in the mine, I took him as my father’ (interview, Solomon).
Solomon recalled that, ‘the time we met there was already an issue at Impala, a [unprotected] strike. So, we started realising that the reason why Impala has gone on strike, is because of these same kind of issues. That's how we started to engage each other.’ They agreed that the workers needed to decide for themselves the way forward. They discussed issues:
Like it seems as if the company is robbing us here and there. The [risk] allowance we won't get it when we are working at night shift, but it is the same as if you are working during the day … the area that you are working at, it's risky. (Interview, Solomon)
More specifically, they each began to hold meetings with other workers in their respective zones:
What we have done is that each and every day when we go underground we had meetings. So in our workplace we started explaining to our crews, we were working with different crews me and … [Anele] I was working the other section, so I had a meeting with people I'm working with, trying to address this issue from Impala about how things should go [here at Amplats]. And then before we go underground now [for meetings], after we realise we have more than twenty people, we had a meeting on the surface so other people can come. We addressed these issues and [then] we would go check the development until June [came]. (Interview, Solomon)
A small group of workers, including Solomon, Anele and then also Sipho, who had come to Amplats in 2011, began organising. Their ad hoc group which was uniting workers could not yet be called a workers’ committee. At first it was difficult for these leaders as some of the workers believed they would be fired. Others, though, thought they should engage with management around wage demands. The numbers began to increase to around 10 or 20. During the months of April and May, they had simultaneously begun a process of recruiting for the new union (AMCU) and organising independently from the NUM. In June, when the number became large enough, more than 20, they began meeting above ground with the day shift and other occupations. Sipho, another worker leader, recalled that:
Most of the employees were interested in what we usually talked about … Every day, every night we could discuss, what can we make so that we put our demands in front of the company that all of us, all of the classes underground, we want not only RDOs or winch operators, but all of us, we want at least our salaries to be adjusted.10
Solomon and Anele decided that in order to expand their organising efforts, they needed to draft a memorandum. It included the demand for a basic salary of R10,000, slightly more than the R9000 that was demanded at Impala. When they included a range of allowances, the total came to R16,070. The two workers met together: ‘We draft this sitting in the changing room … when we went to the mass of plus [or] minus fifty people to a hundred, we were on the surface now.’ It was concluded that:
it was simpler because everyone understood it … they adopted it and said it is good, it makes sense when we demand safety allowance, bonuses, basic, and I think it's six or seven demands to make it [R]16,070. (Interview, Solomon)
In June, workers became aware that RDOs had been given an increase of R750 per month and then promised an additional R250. In March 2012, RDO meetings began to take place, as had also happened since the late 1980s at least at Frank shaft at Amplats (see Steward 2013). Winch drivers were then asking themselves:
why [did] the company decide to increase the salary of the RDOs only and not us as the winch operators and other levels such as stop tillers and so forth? All of us [should] … get an increase. So [in his view] that's where everything started. (Interview, Sipho)
Solomon and Anele discussed this issue together and, in addition to the other issues that they were using to prompt workers to join them, Anele said, ‘we must sell this to the workers’ (interview, Anele). At one shaft, Khuseleka, the process of organising was beginning to gain traction. As one of the strike leaders later stated, ‘We can look out for our own interests. Even AMCU is not part of this action because we do not need unions to represent us’ (in Magome 2012). Soon, hundreds of people from both night and day shifts and from all occupations and unions at Khuseleka were beginning to attend meetings. ‘As we go forward’, according to Sipho:
… everyone was seeing that we are serious, uh, it was around July … where we decided that we are not talking unions here. We are talking salary adjustment. So each and every employee that feels that we are getting a little, let's come together, we draft a memorandum and then we go and submit to management. So we went on the 12[th] of July to submit the memorandum of 16,070 rand … the whole of Khuseleka mine, known as Jabula, was there to support us. And that's where we were chosen by the workers [they said], ‘since you are in the front line, you might as well lead us for the meanwhile.’ Because it might happen that if the management refuse then we choose to go on strike. And if we go on strike, when we come back we need to have new leaders because the current leaders [in NUM], they [are] failing us. (Interview, Sipho)
A top seven was chosen. Within that seven, Solomon and Anele were the primary leadership and it was they who met with management at Khuseleka on at least three occasions between June and July. Solomon described a situation whereby ‘the company nominate[d] me and [Anele] … I was [also] leading them [workers] on the other end.’ He concluded that he was essentially a ‘middleman’ between the workers and company. The company took the contact details of the two workers and told them, ‘as from today, if there are any grievances from the workers … we know you, we will negotiate with you’ (interview, Solomon). Solomon kept a diary of their meetings, explaining that ‘we were following all these processes in order that if the company wants to fire us, we show them [the records] that we have followed the proper procedures’ (interview, Solomon). The numbers attending the mass meetings outside Jabula hostel swelled and virtually all the workers at the shaft were now committed to the common memorandum.
The few informal leaders had thought that it was time to go on strike sometime in August, but ‘the workers stop us and say “Don't! Let's follow the procedures”’ (interview, Solomon). Neither the unions nor the management would respond to the workers’ grievances, so they decided to take the case to the Commission for Conciliation, Mediation and Arbitration (CCMA) on 25 August.11 By early September they had not obtained a response and it was then that workers at a mass meeting, alongside the workers’ committee, decided to plan for strike action with the rest of the shafts at Amplats:
So we agree that from the 8[th] to the 11[th September] we go on strike, all of us. During these three days we used to drive all the operations of Anglo trying to make sure that the strike starts on the same day. (Interview, Solomon)
Sipho recalls that the workers’ committee leadership told the people at Amplats’ Jabula shaft, ‘we can't make it alone. Let's rather go to other operations at Anglo … to give the memorandum to them.' At first it was difficult for other workers to join as they were afraid of being retrenched, but within the next few days, they were able to identify leaders in various shafts that bought into their programme. Solomon explains:
… we move with them group by group to the shafts … We make it simple that you here in other shafts [besides Khuseleka] will have to elect leaders here that will lead you. Because we cannot afford to come and hold the meeting here and go back to Jabula [the hostel near Khuseleka] on a daily basis. (Interview, Solomon)
The idea of the strike now began to spread to others in the region. Thembelani, Siphumelele and Khomanani shafts each had already been organising separately, and Khuseleka would become the spark that ignited them. Their concerns and demands had also been consistently ignored by the branches of the NUM in their shafts, thus making the idea of a general strike very attractive.
According to one of the members of the top seven, Khuseleka then was able to identify the leaders of these shafts who responded positively, saying that Khuseleka is ‘coming with something that we already wanted’ (interview, mineworker). They asked leaders of Khuseleka to:
please come to our shafts and talk to our employees about how far you are or what do you want and tell them this is what we also want to do. And so, how can we strengthen ourselves in order to make sure that [we win]. (Interview, mineworker)
Each of the shafts held a final mass meeting and then on the 11th, the nightshift at Khuseleka, and the 12th generally, no one was to go underground.
Unfinished business: the rocky transition from the workers’ committee to AMCU
On 18 November 2012, the strike at Amplats finally ended.12 According to a reliable source, every worker at Amplats was given a staggering bonus of R24,000 upon their return to work. Effectively, the committee had gained legitimacy in the eyes of both the employer and workers who eventually chose AMCU as their union. They then elected virtually the entire workers’ committee into office-bearing positions in early 2013. There was a rocky and incomplete transition from the workers’ committee to AMCU which we can now explore. On the one hand, it was characterised in part by the oligarchic tendencies of the union's NEC. On the other hand, it confirms that the union's dominance in Amplats results from the insurgency which was shaped and driven by the worker leaders during the unprotected strike in 2012.
Management at Amplats needed a structure with which to engage that was legitimate in the eyes of the workers, or in their view risk anarchy. The company therefore issued ‘interim access’ to what they called the ‘interim committee’, made up of worker committee members (Anglo American, no date, 1). The agreement presented to the workers’ committee stated that ‘the granting of the interim access to the interim Committee as set out in these guidelines was a precondition to the return to work by the supporters of the Interim Committee.’ Management made clear, however, that they were not granting organisational rights, but merely providing access to ‘office facilities and basic amenities.’ The agreement concluded by insisting that the committee would ‘have to be registered as a trade union in accordance with the provisions of the Labour Relations Act of 1995 (“the LRA”) prior to it being eligible for recognition at the operations of Rustenburg Platinum Mines’ (Ibid., 1–2). All Rustenburg Amplats units (mine shafts) were thereby granted up to three part-time representatives and one full-time. This also included the Northam units, Amandelbult and Union Mines. The purpose of these representatives, from the perspective of management, was to resolve conflicts in the workplace in order to ensure productivity. The interim access was to end on 31 December 2012 (Ibid., 1–2).
Based on his own participation in the events at the time, an exceptional organiser in Rustenburg described a situation whereby:
Those offices were run by the workers’ committee … Workers’ committees were afforded transport. If they wanted to meet, they would just request to the relevant department to say we want to go to a meeting … so the comrades from Rustenburg [from different shafts within the committee] all of them they would get a 22-seater [a large minibus] to transport them. They were given that right to go and meet because management understood that these are the committees who represent the workers [and] that keep them [workers] productive. [They] represented the workers without any unions. (Interview, anonymous)
Both management and the committees were aware that eventually the workers would need to join a union. At this stage, workers were debating which union to collectively join. One thing they knew was that they were not going back to the NUM. The leaders of the workers’ committee engaged each other and the rank and file about the way forward. Anele further elucidated what had transpired at the end of the 2012 strike prior to the transition to the new union:
If we agree to go back to work there must be some conditions … We must not show that we are defeated … by going back to work. We were thinking for the masses that are at our back and secondly we saw the difficulties of fighting the employer while you're outside [of the formal bargaining structures]. Because when you fight the employer while you are outside, you are calling the police. We saw that at Marikana. Their people were killed so now in order to prevent what happened at Marikana, so we decided to go back … so that when we are inside [we can then approach management under the auspices of a union]. (Interview, Anele)13
The workers’ committees, operating in direct consultation with the rank and file, without engaging an institutional structure, had limitations at Amplats. It was clear that they had done an impressive job but the committees believed they needed to be inside a formal institution in order to continue to press for changes. One worker who became a shop steward of AMCU joined the union after the strike. He expressed a sharp contrast between the approach of AMCU and the NUM, of which he had been a member:
I had no organisation [during and directly after the strike] and I was in the field. And we were just there as workers only and NUM was not there. And when we came back from the field, then AMCU promised us that they will do what we wanted. And he [Mathunjwa] also said that the money that we want, he will not say [to us later] that it is too much … but he will go to the employer and talk to them about it and he will try to get me that money exactly. And NUM will tell you that the money that you want is too much before even going to the employer. And then AMCU came and told us, ‘no, I will take your mandate as it is and take it to the employer without changing it.’ And [now] they [AMCU] are using our mandate as workers. (Interview, mineworker)
AMCU had not only emerged as a result of the insurgency of the committees, but it had also promised, to a certain extent, to engage rank-and-file workers before making decisions in wage negotiations.
Though AMCU did not have a stronghold in any of the shafts at Amplats in Rustenburg, it had gained members through a man popularly known as Steve. He became a regional organiser of AMCU after he was suspended as the NUM branch chairperson at Karee shaft in Lonmin. Throughout the strike of 2012, the union was gaining more and more momentum (even if it was not reflected in stop-orders) since it was the only union which seemed to the majority to represent the interests of the working class. In January 2013, Amplats announced that it would retrench 14,000 workers, arguably in order to tame the workers in the most militant shaft where the 2012 strike emerged – Khuseleka. Later in the year, the company backed down and planned to cut only 4800 jobs, laying off 3300 workers (The Mineworker 2013). On 27 September Amplats workers went on strike to demand ‘zero’ retrenchments. Throughout 2013, the offensive struggle for a living wage was therefore put on hold as the workers became involved in a defensive struggle against retrenchments. Furthermore, a broad committee made up of about 47 people, representing each shaft, was transitioning into AMCU and seeking to put it into power as the majority union.
On 3 March 2013, Amplats officially signed a recognition agreement with AMCU. Lonmin and Impala had already signed with AMCU at this point. The spokesperson for Amplats, Mpumi Sithole, stated at a press conference that 40% of the 60,000 workers at Amplats were members of AMCU (“Amplats Signs AMCU Recognition Pact.” 2013). Mathunjwa concluded, ‘We have full organisation rights and will now be able to implement formal structures and appoint shop stewards at Amplats’ (Ibid.).
Prior to gaining official recognition, the AMCU NEC was already holding meetings with the workers’ committee and other leaders of Amplats in order to collectively plot a way forward. As early as February 2013, a former leader of the broader committee of about 47, who became an AMCU branch member of Khomanani shaft, could sense the way in which the victories of the workers’ committee could be undermined when they became part of AMCU. During a meeting on 24 February with AMCU representatives and the various leaders at Amplats, he asked Mathunjwa what could be done in order to keep the workers’ committee alive. Mathunjwa's response was straight and to the point when he stated that he was ‘not interested [in the] workers’ committee and their issues. They [the workers and AMCU] are here in the name of AMCU. They [AMCU] are not going to discuss anything concerning the workers’ committee’ (interview, branch leader at Khomanani). For Mathunjwa, the committee at Amplats was temporary and therefore no longer necessary. Furthermore, he was aware of the fact that it was an independent organisation which could potentially challenge the authority of his established union. Instead of viewing the committee as a complementary force, Mathunjwa considered it a threat and therefore aimed to destroy it as an independent factor.
It became apparent to many worker committee members that AMCU was now in the driver's seat. Still, others continued to see the need for a committee that was autonomous from AMCU and which was therefore able to represent the interests of the workers outside of the union's ‘official’ structures. One branch leader, for example, conceded in mid 2013 that ‘once now we fall under AMCU this means that now the workers' committee is no longer there … but at the moment we are discussing as a workers' committee … [how] to do things on our own’ (interview with Khuseleka branch leader). Solomon also reflected that the committee no longer continued to meet officially, but that the branch members were well respected as worker committee leaders in the respective shafts and they therefore had a responsibility to act as watchdogs:
So this is the reason of workers’ committee that after working hours let's go and meet and observe AMCU. [To ensure] that it's doing what we want. Because another thing we said, we took almost like 30 years to form a workers’ committee. We can note that when NUM started operating there was no [prominent] workers’ committee. So we sat and said from the history of 30 years, we can't just kill it in one day. Because it doesn't mean that we trust AMCU [to act in the interest of the workers]. (Interview, Solomon)
Offering a balanced account of the transition to the formalised union, he said that other leaders ‘took their heart completely and give it to AMCU’ (interview, Solomon). He estimated that of the 47 people who were part of the committee during the strike of 2012, by mid 2013 two or three people from each shaft – about 15 in total – remained part of the committee in mid 2013. They met about every two weeks depending upon the circumstances.
The relationship between the AMCU NEC and branch committees became steadily antagonistic and by late 2013 individuals from the branches – and even the North West region as whole – began to challenge the leadership style of the national structure. A letter directed to the head office in early November 2013 claimed that the AMCU regional organiser of the North West was unfairly dismissed ‘without involving other mines including ANGLO platinum … [this] is where you as [the] national office created disrepute within [the] organization without proper consultation’ (AMCU Limpopo and Northwest 2013). They further highlighted several issues including a preference for bottom-up rather than top-down approaches to decision making within AMCU. They also requested the minutes of the NEC meeting which unilaterally resolved to suspend the organiser as well as a performance sheet which outlined his failure to fulfil his duties as an organiser in the region. The letter concluded by questioning the integrity of the union and also by beginning to make a comparison to its arch-rival:
But if AMCU is a one-man show like NUM, please explain to us directly so that we know there is no need of [the] constitutional structure to exist, it means that AMCU operates like a business whereby there [is] no need for consultation [but] only instruction … It seems as if the union does not have [a] political agenda to drive [the] union forward but [that] it has a discrimination/personal agenda. (AMCU Limpopo and Northwest 2013)
At a meeting at one of the shafts that Mathunjwa visiting to engage with the branches, one branch leader described a situation whereby the chairperson of the shaft raised his hand to request some assistance from the national office and Mathunjwa began to shout. He proceeded to ask, ‘you never raised your hands with Zokwane while he was speaking, [so] why are you doing that to me?’ (interview, mineworker). Key branch leaders of AMCU felt disgruntled by Mathunjwa in particular and a new union called Workers Association Union was formed on 3 March 2014. Although there is not space to provide detail, the union has virtually no support. Nevertheless, it represents the ongoing inter-union rivalry in the platinum belt.
At a mass meeting at Olympia Stadium in Rustenburg on 19 January 2014, Joseph Mathunjwa made what was to become a historic move when he announced that the workers were going on strike. The strike began in earnest on Thursday 24 January 2014 as a stand-off between AMCU workers and the three largest platinum mines in the world. The workers were united from the start of the strike, seeking a living wage of R12,500. The vast majority of mineworkers at Amplats, Impala and Lonmin (between 60,000 to 80,000) downed tools in what amounted to the longest strike in South African mining history (five months) as they negotiated for substantial wage deals of about a 20% increase per year for the following three years.
Battling their own hunger and that of their families, workers were unwavering in their commitment to a ‘living wage’, first under the auspices of the workers’ committee (in 2012) and later as rank-and-file members of AMCU (in 2014). Perhaps more important than the fact that the three companies were united in their common demand under the protection of their union, was the radical democratic practice of AMCU which was maintained during the 2014 strike. Though Mathunjwa undermined branch leaders who offered conflicting opinions to his own, the union simply did not make decisions without consulting the mass of workers.
Indeed, the oligarchic tendencies of the union leadership at the national level in relation to branch leaders only tell part of the story. Workers took democratic control of the great 2014 strike under the auspices of the union as they had done formerly under their worker committees. The underlying strength of the workers remained constant over time, thus blurring the division between the industrial bargaining process under the leadership of AMCU and the informal worker committee's leadership.
Conclusion
As Gouldner (1955) indicated, contrary to Michels (1915), oligarchic tendencies within trade unions are only possible to the extent that the rank and file as a collective does not exhibit democratic practices and ideals. Put another way, power within trade unions is something which is in flux and can never be strictly monolithic. It is perhaps of no surprise that the union's strength, and indeed its politics, are defined both by rank-and-file members and by those with official positions. AMCU is the product of the militant labour struggles, in this case the relatively short-lived but extremely potent workers’ committees, from which it emerged. Leaders of the workers’ committees at Amplats, Implats and Lonmin organised independently from unions and embarked on unprotected strikes. When these strikes ended, worker leaders believed they needed a union to represent them and they chose AMCU. The union began championing the radical wage demand of R12,500 after rank-and-file workers had died on the mountain in Marikana waiting for their employer to meet that wage demand (see Alexander et al. 2012). Mathunjwa then took the demand of R12,500 and made it his union's pillar. In a sense, AMCU and Mathunjwa's rise to prominence in the platinum belt has been drawn out of the blood of the 34 mineworkers killed during the Marikana massacre. Hence one prominent AMCU T-shirt proudly worn by mineworkers in the platinum belt reads, ‘Never Forget: We Died for a Living Wage.'
AMCU is neither the saviour nor the enemy of the working class. Rather, the union as an entity in itself is riding on the wave of the insurgent fervour of the rank and file. Like the NUM's form of organising during the 1980s, it too – in this new era – has had to struggle to keep up with the rising tide of working-class power. The article has demonstrated that the fine line between what is autonomous (workers’ committee) and what is formal (trade union) may become blurred in practice. The fact that the dominant view held by both the state and society at large is that unprotected strikes are ‘illegal’ (or anarchic) leads one to delegitimise any organisation which is associated with this behaviour. However, the committee did not need to sign a paper granting it formal collective bargaining rights in order to gain management legitimacy. Management negotiated directly with leaders of the committee during and after the strike of 2012. And the committee had the trust of workers on whose behalf they negotiated with management.
With the exception of some of the former members of the committee, most workers do not see a firm break with the politics of the workers’ committees but rather view AMCU (in particular its face and spokesperson – Mathunjwa) as the embodiment of their struggle. That has been reinforced since the mineworkers won major increases after the 2014 strike. The concept of Insurgent Trade Unionism (ITU) assists in providing an explanation of the relationship between the past when workers went on unprotected strikes with their committee at the helm (2012), and the present when workers went on protected strikes under the AMCU banner (2014). We are currently not witnessing an ordinary trade union, but one that came into power following a mass upheaval. The concept of ITU both highlights the diversity of existing trade union experiences and practices and also demonstrates how they change over time as a result of shifting structural circumstances. Most importantly, it reveals the ways in which trade unions may be driven from below by the rank and file's collective and, in this case insurgent, agency.